• Sherlock, Matthew
• Williams, Alan
• Filisetti, Andrew

Abstract

Scientific net sampling on the seafloor using a variety of trawls and sleds provides valuable information on biota and habitats to biologists and fisheries managers. Specifically, the species composition and stock densities of fishes and invertebrate mega-benthos are routinely derived from net sampling operations to determine stock health and provide indications of the potentially negative effects of environmental changes and human activity on seafloor (benthic) ecosystems . Relatively large net sampling tools such as fish trawls, beam trawls and benthic sleds can be used as platforms to carry imaging technologies to acquire vision of the seafloor, for example on the headline of a fish trawl net (Figure 1). Vision in the form of video and photographic images of the seafloor has the potential to provide highly valuable scientific insights into habitat diversity, species distribution, animal behavior and the direct impacts of human activities including fishing and mining.Australian marine science has a long (5 decades) history of using self-contained underwater film cameras and 'third wire' cable-connected video cameras on trawl nets.Whilst successful, the process was very labour intensive, for example requiring camera pressure cases to be opened between deployments to re-load the film camera and on-board processing of films in a photographic dark room. .Newer digital technologies provide the means to more easily capture high quality image data with higher reliability and much improved work flow for post-capture analysis.The trawl camera system described in this paper was designed to provide a self-contained, autonomous platform suited to deployment on a variety of net samplers.It needed to have a maximum operating depth of 4000 meters and capture high resolution digital images of the seafloor at a relatively high capture rate with excellent reliability and long endurance. The choice of digital image capture in preference to video was based largely on practicality. Video capture at typical operating depths would require the use of one or two subsea lights to provide sufficient illumination at a working height of 3 to 5 meters above the seafloor. The battery capacity required to support several hours of video capture has significant consequences for system weight and turnaround time between deployments.Furthermore, it can be impractical at sea to download, store and process large video datasets.The camera system developed comprises a custom designed frame enabling attachment to a particular platform – in this example, the headline of a fish trawl net (Figure 2). Two separate housings each rated to 4000 meters depth are fitted within this frame. The first housing contains the digital camera (Canon M5), custom designed microcontroller based supervisory circuit, pressure transducer and Li-Ion battery pack.The second housing contains the Xenon flash.Additionally, the platform contains a pair of parallel aligned 520 nm green lasers which provide a known reference distance (200 mm) projected on the seafloor in each digital still image.This distance, in conjunction with the known height of the platform off the seafloor, enables the camera field of view to be estimated. To conserve battery power during deep deployments, the system can be configured to power up the camera, flash and lasers at a preset and adjustable depth. The microcontroller commences taking pictures at this depth at a typical rate of one photo every two seconds. At typical tow speeds of <2 m.sec-1, this rate is adequate to provide a mosaic of the seafloor along the center of the tow track. On recovery the system microcontroller shuts down operations once the platform is shallower than the activation depth.Data offload, battery charging and changes to configuration settings are achieved via cable

Topics

• impedance spectroscopy
• activation
• size-exclusion chromatography
• wire
• aligned